Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning

TL;DR

This paper presents a goal-driven approach to training visual question answering and dialog agents using deep reinforcement learning, demonstrating emergent communication and improved performance on real-image datasets.

Contribution

It introduces a cooperative multi-agent framework with end-to-end RL training for visual dialog, including emergent language without supervision and superior results on real datasets.

Findings

Agents develop their own communication protocol in synthetic environments.

RL fine-tuning outperforms supervised learning on real-image datasets.

Agents learn to ask more informative questions, improving team performance.

Abstract

We introduce the first goal-driven training for visual question answering and dialog agents. Specifically, we pose a cooperative 'image guessing' game between two agents -- Qbot and Abot -- who communicate in natural language dialog so that Qbot can select an unseen image from a lineup of images. We use deep reinforcement learning (RL) to learn the policies of these agents end-to-end -- from pixels to multi-agent multi-round dialog to game reward. We demonstrate two experimental results. First, as a 'sanity check' demonstration of pure RL (from scratch), we show results on a synthetic world, where the agents communicate in ungrounded vocabulary, i.e., symbols with no pre-specified meanings (X, Y, Z). We find that two bots invent their own communication protocol and start using certain symbols to ask/answer about certain visual attributes (shape/color/style). Thus, we demonstrate the emergence of grounded language and communication among 'visual' dialog agents with no human supervision. Second, we conduct large-scale real-image experiments on the VisDial dataset, where we pretrain with supervised dialog data and show that the RL 'fine-tuned' agents significantly outperform SL agents. Interestingly, the RL Qbot learns to ask questions that Abot is good at, ultimately resulting in more informative dialog and a better team.